erikb

Although we are just a small part (about 25 staff members) of a department we still get quite a bit of spam. The spam fills up our mailboxes and our log files.

As postmaster for our domain I find the reading of the mail logs difficult and boring. Simply put, it never happens...

Each night, I automatically filter the mail log files into a readable text summary. That works ok but it can still be improved. In the next stage I transform the text summary into a periodic spam table for the last days. The transformation into a periodic spam table is done dynamically, per request.

Here's how to read the entries in the table:

I also extract some extra information, about days with the higest and lowest ratio of spams in relation to non-spam. See the table below.

4 days ago	3 days ago	2 days ago	yesterday
The day with most spam in relation to non-spam	The day with least spam in relation to non-spam	The day with higest number of spam

The color coding is just a classification of the spam ratio.

See the result.